fair machine experiment interface · gsi helmholtzzentrum für schwerionenforschung gmbh ralph j....
TRANSCRIPT
GSI Helmholtzzentrum für Schwerionenforschung GmbH 1Ralph J. Steinhagen, [email protected], 2015-02-24
FAIR Machine Experiment Interface – First Iteration & Proposals –
Ralph J. Steinhagen
With input from: FC2WG, R. Bär, F. Hagenbuck, D. Ondreka, D. Severin
GSI Helmholtzzentrum für Schwerionenforschung GmbH 2Ralph J. Steinhagen, [email protected], 2015-02-24
Outline
● Introduction: New FAIR Acc. Control System Highlights
● Machine-Experiment Interfaces:
a)Accelerator & Beam-Modes, Beam Mode transitions
b)Target Steering, Beam Performance Indicators & Macro-/Micro-Spill-Structure
c) Interface to Machine Protection (MP): Beam Aborts
… with focus on controls & software interfaces
GSI Helmholtzzentrum für Schwerionenforschung GmbH 3Ralph J. Steinhagen, [email protected], 2015-02-24
Fundamental Axiom
An accelerator is more than the sum of its parts:
FAIR-MCR &human-machine interfaces
Control System: LSA, Archiving/Logging, Performance Monitoring
Beam-Based Systems/BI Integration(alignment, feedbacks, …), MCR Applications, ...
Comissioning Procedures & Facility Exploitation
Personnel development & training
ma
gne
ts
inst
rum
en
tatio
n
RF
othe
r H
W s
yste
ms
..
.
SWHW
e.g SIS18
ma
gne
ts
inst
rum
en
tatio
n
RF
othe
r H
W s
yste
ms
..
.
e.g SIS100
ma
gne
ts
inst
rum
en
tatio
n
Str
ipp
er
othe
r H
W s
yste
ms
..
.
HEBTm
agn
ets
inst
rum
en
tatio
n
RF
othe
r H
W s
yste
ms
..
.
e.g. HESR
...
mac
hin
e-to
-mac
hin
eIn
terf
ace
mac
hin
e-to
-mac
hine
Inte
rfac
e
mac
hin
e-to
-mac
hin
eIn
terf
ace
FA
IR U
ser s/E
xperim
en ts
FAIR-MCR (“HKR”)
FC2WG
MPLs/MCs
Beam Commissioning
Sys
tem
Int
egr
atio
n &
Co
ntr
ols
Ha
rdw
are
Co
mm
issi
on
ing/
Dry
-Ru
ns
GSI Helmholtzzentrum für Schwerionenforschung GmbH 4Ralph J. Steinhagen, [email protected], 2015-02-24
Example: FAIR Commissioning Procedureshttps://fair-wiki.gsi.de/FC2WG/BeamCommissioning
GSI Helmholtzzentrum für Schwerionenforschung GmbH 5Ralph J. Steinhagen, [email protected], 2015-02-24
Anatomy of FAIR Acc. Control SystemR. Baer (department head) et al.
LSA-DBLSA
(Settings Mgmt)(Java)
Archiving(Java, C++, tbc.)
short-term(few weeks)
long-Term(years → ∞)
Prop. HW(non-FESA)
Mac
hine
-Pro
tect
ion,
Inte
rlock
s, M
AS
PT
rans
mis
sion
Mon
itorin
g, P
ost-
Mor
tem
Applications(Java)
Applications(Java)
Applications(Java)
Applications(Java)
Applications(Java)
Timing(FESA, HW, C++)
Front-End(FESA, C++)
Front-End(FESA, C++)
actual HW actual HW actual HW
...White Rabbit
Tier 3: Industrial Control & IT(aka. Front-Ends)
Tier 2: Business Layer(Control Logic)
Tier 1: Application/PresentationFAIR Facility Overview
('Page 1')Applications
(Java)
Applications(Java)
Exp. Apps.(Java)
SCADA(for prop. HW,
e.g Cryo.)
LayoutDB
(???
common generic API
JAPC (CMW = ZeroMQ-based)
GSI Helmholtzzentrum für Schwerionenforschung GmbH 6Ralph J. Steinhagen, [email protected], 2015-02-24
Accelerator Controls RetrofittingGSI → FAIR Transition in 2018
2018
GSI Helmholtzzentrum für Schwerionenforschung GmbH 7Ralph J. Steinhagen, [email protected], 2015-02-24
Some FAIR Acc. CO Highlights
● Non-monolithic, distributed accelerator control system
– Standardised open interfaces across accelerator chain
● Full control on accelerator modelling & interdependencies
– Consistency across devices and settings
– Parameter hierarchy (trans. abstraction/separation of: accelerator physics ↔ controls)
● Tracking of accelerator performance & corresponding settings
● Distributed development & expertise (e.g. code shared with CERN)
– Better more efficient code review & medium-/long-term maintenance
– Continuous improvement & evolution to actual accelerators' & experiments' needs
● Better more flexible integration of beam-based diagnostics systems
– possibility of arbitrary (user-driven) functions (+actual vs. reference comparisons) → opens the possibility of powerful semi-/fully automated cycle-to-cycle feedbacks
● trajectory/target steering, orbit control, …
● Easy macro-spill-control,
● ...
GSI Helmholtzzentrum für Schwerionenforschung GmbH 8Ralph J. Steinhagen, [email protected], 2015-02-24
Machine-specific Beam-Based Systems:
● SIS18: multi-turn-Injection (N.B. highly non-trivial,
complex subject), Slow-Extraction (K.O. exciter, spill‑structure, …)
● SIS100: Slow-Extraction (K.O. exciter,
spill‑structure, …), RF Bunch Merging and Compression
● ESR, HESR & CR: Stochastic cooling, Schottky diagnostics, …, tbd.
Generic:
● Remote DAQ of Analog Signals (strong impact on HKR migration/operation!)
● Facility-wide fixed-displays, facility & Machine Status (“Page One”)
● context-based monitoring of controls and accelerator Infrastructure,
● … “the sky is the limit”
Appetizer: – LSA enablingBeam-based cycle-to-cycle feedbacks
Generic Beam Control (focus on use-case)
1. Transmission Monitoring System (R. Steinhagen, FC2WG Meeting #6)
2. Orbit Control (work in progress)
3. Trajectory Control (threading, inj./extr., targets)
4. Q/Q'(') Diagnostics & Control
5. TL&Ring Optics Measurement + Control (LOCO, AC-dipole techniques etc.,)
6. RF Capture and (later) RF gymnastics
7. Longitudinal Emittance Measurement
8. Transverse emittance measurement
9. Transverse and longitudinal feedbacks
Bre
ad-a
nd-B
utte
rsy
stem
s fo
r O
P~
idea
lly f
or S
IS18
res
tart
GSI Helmholtzzentrum für Schwerionenforschung GmbH 9Ralph J. Steinhagen, [email protected], 2015-03-09
LSA
PC-GatewaysPC-GatewaysPC-GatewaysMonitor-FrontendMonitor-Frontend
...
BPM Concentrator
Orbit/Trajectory(JAPC)
Monitor-Frontend
Ethernet UDP/IP/JAPC
beam response
Abstraction Layer/Testbed
Cycle-to-Cycle FB GU (Orbit Representation, etc.)
SurfaceTunnel
...beam instrument
LSA Trims(JAPC)
corrector magnets
m x n x
Middle-TierFront-End Tier
Orbit/Trajectory Controller
User/GUI-LevelData Presentation
LSA Trims
Reference optics, Reference orbit
GUI/Expert Applications (e.g. “YASP”)
simulation
Example: LSA-based Orbit Control B. Schlei et al. (being prepared)
GSI Helmholtzzentrum für Schwerionenforschung GmbH 10Ralph J. Steinhagen, [email protected], 2015-02-24
Example: LSA-based Macro-Spill & Harmonics Control I/II
courtesy V. Kain, J. Wenninger, CERN
blue: reference red: ring-DCCT
Autopilot: low gain
from SISMODI:
to LSA-based cycle-to-cycle feedback:
GSI Helmholtzzentrum für Schwerionenforschung GmbH 11Ralph J. Steinhagen, [email protected], 2015-02-24
Example: LSA-based Macro-Spill & Harmonics Control II/II
existing tune reference change
original tuneproposed tune
proposed tune correction
The Autospill interfaces via the standard LSA trim client (API) → trim history, propagation to other parameter… courtesy V. Kain, J. Wenninger, CERN
GSI Helmholtzzentrum für Schwerionenforschung GmbH 12Ralph J. Steinhagen, [email protected], 2015-02-24
LSA-based Cycle-to-Cycle Feedbacks
courtesy V. Kain, J. Wenninger, CERN
Generic LSA trim interface:
… open to all accelerator parameter that fulfil basic control theory criteria:
● Stability: “parameter should be ~ reproducible from fill-to-fill … “
● Observability: “... need to be able to measure it reliability (noise, …), ...” → your Exp. feedback desired: the better the info/quality of parameter deviations the better the FB!
● Controllability: “need affine (but not necessarily linear dependence between observable effect and control actuator, ...”
GSI Helmholtzzentrum für Schwerionenforschung GmbH 13Ralph J. Steinhagen, [email protected], 2015-02-24
FAIR Control System Documentation
Good entry point:
https://www-acc.gsi.de/wiki/
Specifically, some more meat in:
● Glossary: https://www-acc.gsi.de/wiki/FAIR/FAIRGlossar
● Timing: https://www-acc.gsi.de/wiki/Timing/
● FESA: https://www-acc.gsi.de/wiki/FESA
– https://www-acc.gsi.de/wiki/FESA/WhatIsFESA
● LSA: https://www-acc.gsi.de/wiki/Applications/LsaMainPage
– https://www-acc.gsi.de/wiki/Applications/LsaPresentationsAndPublications
– https://www-acc.gsi.de/wiki/Applications/LsaFrequentlyAskedQuestions
– https://edms.cern.ch/ui/#!master/navigator/document?D:1935804008:1935804008:subDocs
● Applications: https://www-acc.gsi.de/wiki/Applications/
GSI Helmholtzzentrum für Schwerionenforschung GmbH 14Ralph J. Steinhagen, [email protected], 2015-03-09
Machine-Experiment Interfacesa) Accelerator and Beam Modes
GSI Helmholtzzentrum für Schwerionenforschung GmbH 15Ralph J. Steinhagen, [email protected], 2015-03-09
Accelerator & Beam Modes
● Purpose:– Communication of intended accelerator operation to experiments, FAIR and wider community
● what to expect and when, beam time performance tracking & analysis
– Conditioning of control sub-system responses● e.g. logging, interlocks, management of critical settings (MCS & RBAC), machine sequencer, access system, …● associated rules of what is allowed, when, where etc. e.g.:
– Limit parameter changes during data taking – aka. 'Stable Beams'/'Production Runs'– No high-intensity beam injected into an 'empty' machine– Block certain operation during unsafe mode of operation
● Main modes:
1) Accelerator (Machine) Modes● covering rule sets outside of beam operation● defined per accelerator/transfer-line segment
2) Beam Modes● covering rule sets during beam operation● defined per accelerator/transfer-line segment and beam-production-chain
● proposal: extend this concepts also to experiment targets– required for safe primary-beam intensity ramp-up & OP-Exp. Hand-shake (see later)
– more fine-grained options for facility availability, performance tracking & analysis
copy on:https://fair-wiki.gsi.de/FC2WG/
GSI Helmholtzzentrum für Schwerionenforschung GmbH 16Ralph J. Steinhagen, [email protected], 2015-03-09
FAIR Accelerator Modes
Follows annual life-cycle of accelerator facility ● Operation without Beam:
– SHUTDOWN
● could imply possibility of open/controlled access or no powering – COOLDOWN (SIS100, SFRS)
● typ. 2-3 weeks, limited/no access● need to distinguish between a 'warm' and cold' shutdown?
– BAKE-OUT (SIS18, HEBT, …) – similar to cool-down
– WARM-UP (SIS100, SFRS)
– RECOVERY (SIS100, SFRS)
● after quench, partial vacuum loss, typ. few hours - day● includes e.g. periodic magnet CYCLING to stabilise hysteresis
– MACHINE-CHECKOUT
● operations tests without beam in view of beam operation● (e.g. power converter calibration, magnet patrol, etc.)● done once after a long shutdown, typ. few weeks before beam operation
– ACCESS (during beam operation periods)
● controlled access for specific tasks only (signature by MCs & OP)● Operation with Beam:
– BEAM SETUP or MACHINE SETUP● focus on initial/re-commissioning, machine setup after long shut-down + OP training
– PHYSICS
– MACHINE DEVELOPMENT● focus on accelerator/beam physics aspects
– MACHINE TEST (during beam operation periods)● controls, RF, new front-end, ... tests w/o beam + OP training
● Ad-hoc during beam operation but not 'Physics' nor 'MD'
describe main aimof machine operation
+info & Accounting
type modes
operation without beam(part of shut-down coordination)
GSI Helmholtzzentrum für Schwerionenforschung GmbH 17Ralph J. Steinhagen, [email protected], 2015-03-09
Example: GSI Beam Scheduleaka. “Strahlzeitplan”
Accelerator Modes are not new!already practise in normal operation
→ plan to make CS also aware of them→ allows to define rules, statistics, ...
GSI Helmholtzzentrum für Schwerionenforschung GmbH 18Ralph J. Steinhagen, [email protected], 2015-03-09
FAIR Beam Modes – State Diagram
Post-Mortem/Beam Dump
Recovery:No Beam
No Beam
Pilot Beam
Intensity Ramp-Up
Adjust
Stable Beams/Production
cool down + cycling aftermagnet quench or main PS failureN.B. beam mode = machine mode
Here'd be Happinessproducing physics beamsmost settings locked-down
for low intensitybasic accelerator setupinjection->extractiontypically with (but not limited to)low setup intensities (SBF=true)
normal operational patherror/fault caselow-intensity
always start here:
mag. cycles onlye.g. RF conditioning
verification of machine-protection functionalityMinor adjustment of intensity related effects (e.g. ∆Q(intensity))
tune beam parameters (within limits) to suit the experiments needs/performance
“handshake”
N.B.:1) omitted arrows to 'No Beam'/'Pilot Beam' for
better visibility (always possible)2) modes follow existing normal setup routine, initial
transition acknowledged by operator, subsequent driven automatically by sequencer
GSI Helmholtzzentrum für Schwerionenforschung GmbH 19Ralph J. Steinhagen, [email protected], 2015-03-09
● Follows life-cycle of beam setup and production – also not new, de-facto how– NO BEAM
● prevent/stop beam being injected by design (↔ mode)
– PILOT BEAM● Establishing main machine parameters: injection steering, RF capture, ramp, orbit, Q/Q', optics checks, extraction
– typically done with low setup intensities
– INTENSITY RAMP-UP● beam parameter tuning and checks related to increasing intensities (e.g. slow extraction, space-charge, etc.)● check of interlocks & machine protection functionalities (limited user group only ↔ special RBAC role for e.g. interlock settings)
– ADJUST● Perform actions/change minor beam parameters (“minor” needs to be defined)● tune beam parameters (within limits) to suit the experiments needs/performance
– STABLE BEAMS (beam production for physics)● main intend is to deliver stable beam to experiments● tbd: very limited machine tuning
– BEAM-DUMP or POST-MORTEM● response to quench, MP action, or other action that needs to be analysed before one can continue with normal operation
– RECOVERY● recovering from severe post-mortem, essentially includes 'CYCLING'
● For storage rings (essentially only ESR & HESR)– Do we need INJECTION/ACCUMULATION & RAMP modes?
FAIR Beam Modes
Beam Modes are not new!already practise in normal operation
→ plan to make CS also aware of them→ allows to define rules, statistics, ...
GSI Helmholtzzentrum für Schwerionenforschung GmbH 20Ralph J. Steinhagen, [email protected], 2015-03-09
Additional Actual States
● Beam Presence Flag (BPF) – indicates that cycle/settings have been validated with Pilot- or Physics-Beam in the recent past (< days, tbd.)
– main usage: prevent high-intensity injections into an 'empty' machine with new untested magnetic settings or modified machine conditions
– defined per accelerator or transfer-line segment (where necessary)
● Setup Beam Flag (SBF) – indicates beam used to setup the beam production chain (typically low-intensity)
– defined per accelerator or transfer-line segment (where necessary)
– SBF provides flexibility of masking interlocks during setup (e.g. MWPC/screen protection)
– Used to enforce interlocks with high-intensity (primary) beam
● Injection/Extraction Permit – indicates if subsequent accelerator chain is ready (safe) to receive beam (→ fast beam aborts, discussed later)
GSI Helmholtzzentrum für Schwerionenforschung GmbH 21Ralph J. Steinhagen, [email protected], 2015-03-09
Machine-Experiment Interfacesb) Target Steering & Spill-Structure
courtesy C. Bert, A. Constantinescu, D. Ondreka, M. Kirk et. al.
J.Pietraszko, HIC4FAIR Workshop, Darmstadt, February 26 ,2016
day 096
day 097
day 098
day 099
day 100
day 101
day 102
0.6 mm
1.2 mm
1.2 mm
1.2 mm
1.8 mm
0.9 mm
0.9 mm
0.9 mm 0.9 mm
0.3 mm
0.3 mm
0.3 mm
day 103
day 104
day 105
day 106
day 107
Beam position stability – day-wise
time time time
GSI Helmholtzzentrum für Schwerionenforschung GmbH 23Ralph J. Steinhagen, [email protected], 2015-03-09
SIS100 → p-Target Steering AnalysisTyp. trajectory for 0.3 mm r.m.s. misalignment (3σ cut)
● Common control (theory) questions:– Stability → expected orbit perturbations (quads/groundmotion, septa, b1 errors, ...)
– Observability → number/position of BPMs
– Controllability → number/position of steerer→ points look OK to first order, but wouldn't be able to hit the target without any (target) beam diagnostics
raw orbitafter correction
raw orbitafter correction
horizontal vertical
GSI Helmholtzzentrum für Schwerionenforschung GmbH 24Ralph J. Steinhagen, [email protected], 2015-03-09
SIS100 → p-Target Steering AnalysisZoom
horizontal vertical
● Trajectory in transfer line OK (< ~ 3 mm)● Larger deviations at p-Target: ± 2 mm
– Needs some interface definition (BPM near target)
GSI Helmholtzzentrum für Schwerionenforschung GmbH 25Ralph J. Steinhagen, [email protected], 2015-03-09
SIS100 → p-Target Steering AnalysisEffect of additional BPM near Target
horizontal
GSI Helmholtzzentrum für Schwerionenforschung GmbH 26Ralph J. Steinhagen, [email protected], 2015-03-09
SIS100 → p-Target Steering AnalysisAdd. Target BPM + shifted BPM before triplet
horizontal
GSI Helmholtzzentrum für Schwerionenforschung GmbH 27Ralph J. Steinhagen, [email protected], 2015-03-09
CERN-AB-2005-087 – Ground Motion Modelapplied to SIS100 → p-Target
● Ground motion possibly more dominant for p-Target– may need a full check/realign transfer line at least every ~ 10 h
– Very slow drifts: steering bandwidth on the scale of few minutes seems sufficient
<x2>BPM
≈ 10 · <Δx2>quads
<y2>BPM
≈ 1.6 · <Δy2>quads
GSI Helmholtzzentrum für Schwerionenforschung GmbH 28Ralph J. Steinhagen, [email protected], 2015-03-09
HEBT Beam Diagnostics – CBMF. Hagenbuck & B. Walasek-Höhne (BI)
ExperimentHEBT
GSI Helmholtzzentrum für Schwerionenforschung GmbH 29Ralph J. Steinhagen, [email protected], 2015-03-09
HEBT Beam Diagnostics – SuperFRSF. Hagenbuck & B. Walasek-Höhne (BI)
ExperimentHEBT
GSI Helmholtzzentrum für Schwerionenforschung GmbH 30Ralph J. Steinhagen, [email protected], 2015-03-09
Exp. Target Steering Requirements I/II… from an accelerator point of view
● Time dependent beam-on-target position changes → LSA enables very efficient 'cycle-to-cycle feedbacks' that can compensate recurring drifts on compensating this (ms-level spill control)
– Two options: controlling/fixing at the source (angle/position at ring septa), or agnostic with transfer-line steerer magnets (not necessarily best option, but more robust w.r.t. FB control, …)
– Prerequisite: info on actual target beam position needs to be 'known' to the control system
● Few open discussion items w.r.t. Target instrumentation:
● Foreseen? By BI group? → Yes? → OK no problems w.r.t. CO integration
● Position (& profile) at target & after last final-focus magnet!?
– for both setup and nominal physics run (online re-steering)
● Experiment specific diagnostic (assume mostly 'yes')? → benefit of integration via common CO FESA interface
– aim at common standard interface for all exp.
● N.B. could be a simple copy of what already exists
– “if we know (quantitatively) what's bothering you, we can fix it”
OK
? ? ?
12 (3)
J.Pietraszko, HIC4FAIR Workshop, Darmstadt, February 26 ,2016
1ms/div
DAQ Busy signals
Fast reaction trigger signals
DAQ busy and reaction trigger at 1ms scale long periods without reaction trigger
31
1ms/div
GSI Helmholtzzentrum für Schwerionenforschung GmbH 32Ralph J. Steinhagen, [email protected], 2015-03-09
Beam Performance Indicators & (Micro-) Spill-Structure
● Not …
– … a new phenomenon
– … alone in the world with this problem
● But …
– … easier to fix having accurate online information inside control room (via control system: archiving, interlocks, …)
– … the better the closer to the source (where it matters = exp.)
● Using CO standard FESA:
– easier integration
– Maintainability – don't have long-term resources to support short-term hacks
C. Stahl, PhD Thesis, 2015
How does bunching help?
H. Simon, HIC4FAIR WS 2015
Why do particles seem to cluster?
GSI Helmholtzzentrum für Schwerionenforschung GmbH 33Ralph J. Steinhagen, [email protected], 2015-03-09
Exp. Target Steering Requirements II/II… wish list from an accelerator point of view
● Target Steering – FESA integration of Exp. Specific Detectors (if not already based on BI standard, N.B. FESA being a front-end server architecture wrapped around whichever experiment control system)
– x(t), y(t), Δp/p(t) + r.m.s. & status bits → cycle-to-cycle feedbacks (ms-level binning)
– target temperature, BLMs & radiation (interlock) levels (if not part of accelerator infrastucture), ...
– for setup (& online, if available): <σx>,<σy> or if available: σx(t), σy(t), 2D distribution ● at target and directly after last final-focus magnet
● Beam Performance Indicators & (Micro-) Spill-Structure– General Parameters:
● (avg.) bunch length [ns], bunch profile [ns scale] + r.m.s. & status bits● total numbers of particles extracted on target (transmission efficiency)● particles recorded by experiments (accelerator/experiment efficiency), in relation to expected rates (model assumptions:
cross-sections, detector efficiencies) → the real performance indicator 'integrated “useful” luminosity'● beam-induced background information (signal-to-background from an experiment point-of-view → input/discussion
required)– primary beam, secondary beams (up-stream/down-stream generated
– Slow spill structure → cycle-to-cycle feedbacks (binning: 1 ms or 100 us, tbd.)● Binned particle rate dN/dt – total & 'by bunch' (→ bunched beam extraction)● pile-up histogram, spectrum (2D array: 1 FFT per 10% of the spill duration), integrated r.m.s. (from 0 → detector
bandwidth), detector bandwidth
– Fast spill structure → fast in-cycle feedbacks (future option, priority tbd. → disclaimer)● proposal: digital optical Gigabit-link (cables pulled to SIS18/SIS100 K.O. exciter, SW protocol to be defined)
GSI Helmholtzzentrum für Schwerionenforschung GmbH 34Ralph J. Steinhagen, [email protected], 2015-03-09
Interface to Machine Protection SystemBeam Aborts
≠
GSI Helmholtzzentrum für Schwerionenforschung GmbH 35Ralph J. Steinhagen, [email protected], 2015-03-09
Interface to Machine Protection SystemBeam Aborts (R. Baer, C. Omet et al.
1. Programmed, Slow Beam Abort:
– Experiments sets e.g. 'Beam Mode xxx' → 'No Beam' – cycle-to-cycle basis
– Beam request interface to be agreed upon (HW, SW, which beam parameter info, ...)
2. Faster Beam Abort (within-cycle) via timing system (ms-scale)
– requires: FMECA Analysis (soon) & post-mortem (when it occurs) from exp.
3. Fast Beam Abort (FBAS, within cycle) via digital optical link (us-scale)
– requires: FMECA Analysis (soon) & post-mortem (when it occurs) from exp.
● Proof of limited false-positive rate
– Meant for protection against dangerous direct impact of primary beams
– SIS100 FBAS limit on false-positive MP triggers « 1 trigger/day (for nom. Operation))
4. Non-MP-related Spill-Aborts: medical-type-operation/int. dose control
– not part of FAIR base-line ↔ activation issues/where to dump remainder of beam
GSI Helmholtzzentrum für Schwerionenforschung GmbH 36Ralph J. Steinhagen, [email protected], 2015-03-09
SIS18 & SIS100 Beam Dumps I/IIIHard Radiation Limits (C. Omet, RP, et al.)
Dump Official Name Losses*“SIS18 dump” HHD setup-only < 109 @18 Tm 40% (per 10h)
@18 Tm 5%
SIS18 → SIS100 extr.-setup dump
T8DUSD1 setup-only < 109 0.5 Hz max!(no 3 Hz OP with HI!)
90%
1.5·1011 ions/s 200 MeV/u 5%5·1012 protons/s 4 GeV 5%
SIS18 → SIS100 (only @ stripper & collimators)
Tunnel 101 5·1012 ions/s 4.7 GeV/u 2%
SIS100 internal emergency dump
Tunnel 110 (Pt. 9), 1S54SD1
3·1011 U238/s 2.7 GeV/u 10%
2·1013 protons/s 29 GeV 3%
CBM dump T1D1SD1 < 109 ions 0.5 Hz max! 40%
5·1011 ions/s 34 GeV/u 3%
2·1013 protons/s 29 GeV 3%
● Losses* = integrated total losses over 365 days– N.B. typically run < 200 days, not all with high intensity & high energy → gain some factor, – However: limit does not include activation limits (maintainability) and is shared with operational
(unavoidable) losses (injection losses, capture losses, fast/slow extraction inefficiency, halo colimation, …)
GSI Helmholtzzentrum für Schwerionenforschung GmbH 37Ralph J. Steinhagen, [email protected], 2015-03-09
SIS18 & SIS100 Beam Dumps II/IIIHard Radiation Limits (C. Omet, RP, et al.)
Dump Official Name Losses*[...]
SIS100 internal emergency dump
Tunnel 110 (Pt. 9), 1S54SD1
3·1011 U238/s 2.7 GeV/u 10%
2·1013 protons/s 29 GeV 3%[...]
● 1S54SD1 foreseen as protective dump to mitigate beam-related severe machine protection issues but should not be used regularly with nominal beam.
● Losses* are shared between 'regular (un-)controlled losses' & 'exceptional losses'– Regular losses: losses on electro-static septa for slow extraction (expect for SIS100 ~ 10%)
● (N.B. for comparison: SIS18 operation in 2014 → up to 50% nominal losses during slow-extraction)
– Exceptional losses: real machine-protection events + false-positive MP events● Strong impact of false-positive beam dumps
● Two paths:– Need to assess (and minimise) 'false-positive' MP rate → 'Failure Mode, Effects and Criticality Analysis' (FMECA)
for Experiments, similar to what has been done for SIS100 (C. Omet et al.)● Magnets/power converter are straight one-to-one copy of SIS100 results (numbers,● Do not include issues related to SIS100, HEBT (covered in separate FMECA)
– Implement 'dead-time' after false-positive (& also nominal) emergency beam dumps● Intrinsic: acknowledgement (= analysis) of post-mortem event + cool down of dump block (if necessary)
GSI Helmholtzzentrum für Schwerionenforschung GmbH 38Ralph J. Steinhagen, [email protected], 2015-03-09
SIS18 & SIS100 Beam Dumps III/IIIProposal on FBAS & PM Limits:
Dump Official Name Losses*SIS100 internal emergency dump
Tunnel 110 (Pt. 9), 1S54SD1
3·1011 U238/s 2.7 GeV/u 10%
2·1013 protons/s 29 GeV 3%
● Experiment Diversity@FAIR: shouldn't favour single exp. beyond above others (within reasonable limits)
● ALARA:
– Minimise false-positive triggers by design (result of FMECA)
– keep op. losses well below legal limits → permit only a couple of FBAS/day
– limit loss based on average calculated per hour (or at least not more than a day)
● keeps beam-time schedule predictable over a couple of weeks/months, less “dirty” tricks
● Additional dead-time (beyond PM ack. & cool-down) based on actual cycle losses
'Limit :≈ ∫ nlossed(t) [particles] · Ea(t) [GeV] dt / #experiment'
– less penalty for confirmed lower energy or lower intensity losses (→ Transmission Monitoring System, FC2WG #6)
– exponent 'a' to be checked (e.g. LHC (protons) a ≈ 1.7, ions @ FAIR energies?)
● Re-assess limits after first year of operation with real high-intensity beams
GSI Helmholtzzentrum für Schwerionenforschung GmbH 39Ralph J. Steinhagen, [email protected], 2015-03-09
Machine-Experiment Controls InterfaceTentative Summary
● Beam-Modes & Beam Mode transitions– Beam parameter requests ↔ Beam Schedule (aka. “Strahlzeitplan”):
● beam type, charge state, beam rigidity ([Tm]), intensity, spill-rate, … (more?)
– Beam-Mode Transition for Experiments: 'No-Beam' → 'Pilot Beam', ...● Special Machine-Experiment Hand-Shake: Adjust ↔ 'Stable-Beams'
● Target Steering:– x(t), y(t), Δp/p(t) + r.m.s. & status bits→ cycle-to-cycle feedbacks (ms-level binning)
– Target temperature, BLMs & radiation (interlock) levels (if not part of accelerator), ...
– for setup (& online, if available): <σx>,<σy> or if available: σx(t), σy(t), 2D distribution
● Beam Performance Indicators & (Micro-) Spill-Structure– General:
● avg. bunch length [ns], bunch profile [ns scale] + r.m.s. & status bits● total numbers of particles extracted on target (transmission efficiency)● particles recorded by experiments (accelerator/experiment efficiency), in relation to expected rates (model assumptions: cross-sections, detector efficiencies)● beam-induced background information (signal-to-background from an experiment point-of-view → input/discussion required)
– primary beam, secondary beams (up-stream/down-stream generated
– Slow spill structure → cycle-to-cycle feedbacks (binning: 1 ms or 100 us, tbd.)● Binned particle rate dN/dt – total & 'by bunch' (→ bunched beam extraction)● pile-up histogram, spectrum (2D array: 1 FFT per 10% of the spill duration), integrated r.m.s. (from 0 → detector bandwidth), detector bandwidth
– Fast spill structure → fast in-cycle feedbacks● digital optical Gigabit-link (cables pulled to SIS18/SIS100 K.O. exciter, SW protocol to be defined)
● Interface to Machine Protection (MP) → Beam Aborts:– Programmed, slow Beam Abort: Exp. 'Beam Mode xxx' → 'No Beam' – cycle-to-cycle basis
– Faster Beam Abort (within-cycle) via timing system (ms-scale) → post-mortem & FMECA for Experiments
– Fast Beam Abort (FBAS): digital optical link (us-scale) → post-mortem & FMECA for Experiments● protection against dangerous direct impact of primary beams & SIS100 FBAS limit on false-positive MP triggers « 1 trigger/day (for nom. Operation))
– Non-MP-related spill-aborts: medical-type-operation/dose rate control● not part of FAIR base-line ↔ issue of activation/where to dump remainder of beam
Exp. #2 Exp. #3
Exp. (Target) #1
Prio. II Prio. II
Prio. II
prio I
GSI Helmholtzzentrum für Schwerionenforschung GmbH 40Ralph J. Steinhagen, [email protected], 2015-02-24
Comments? Questions?
GSI Helmholtzzentrum für Schwerionenforschung GmbH 41Ralph J. Steinhagen, [email protected], 2015-02-24
Settings Mgmt: Data Supply & ModellingD. Ondreka (AL) et al.
GSI Helmholtzzentrum für Schwerionenforschung GmbH 42Ralph J. Steinhagen, [email protected], 2015-02-24
Front-End-Software-Architecture (FESA)U. Krause et al.
GSI Helmholtzzentrum für Schwerionenforschung GmbH 43Ralph J. Steinhagen, [email protected], 2015-02-24
Machine Timing PerspectiveD. Beck et al.
GSI Helmholtzzentrum für Schwerionenforschung GmbH 44Ralph J. Steinhagen, [email protected], 2015-XX-YY
Accelerator vs. Beam Mode Matrix
concatenation of <accelerator mode>:<beam mode>e.g. 'Shut-down:No Beam', 'Physics:Pilot Beam'
N.B. defined peraccelerator/transfer line segment
N.B. defined peraccelerator/transfer line segment &beam production chain
GSI Helmholtzzentrum für Schwerionenforschung GmbH 45Ralph J. Steinhagen, [email protected], 2015-03-09
Glossary I/III'Mode' & 'Actual States'
Mode:● deliberate user-driven state used to precondition the control system behaviour and responses independent of the
actual accelerator or beam state → 'reference' or 'desired target' of operation (long-term)– formal agreement between accelerator operations and experimental users w.r.t. what to expect
● Tracked by operator (initially) and semi-automated sequencer to follow normal operational sequence– Example 1: 'Shut-Down' →'Cool-Down' → 'Machine Check-Out' – Example 2: … → 'no beam' → 'pilot beam' → 'intensity ramp-up' → 'adjust' → 'stable beams/production for physics' → ...– need to limit number of mutually exclusive and concise modes ↔ overhead of settings generation and their checks
● no real-time requirements
Actual State:● measured current state of the accelerator/beam (short-term)
– perviates accelerator & beam mode definition & equally used as a ad-hoc/post condition– Examples: Beam-Presence-Flag (BPF), Setup-Beam-Flag (SBF), Injection & Extraction Permit (MP interlock states)
● real-time requirements
● Examples: – 'No Beam' beam mode declares intend (as an agreement) that there will be no beam in the machine– 'Beam Presence Flag' is measured actual state whether there is (/was) beam in the machine or not
● N.B. obviously a 'NO BEAM' beam mode & 'BPF=true' should lead to an interlock
● Shouldn't mix 'modes' with 'actual states' to prevent circular dependencies, priority/causality inversions
GSI Helmholtzzentrum für Schwerionenforschung GmbH 46Ralph J. Steinhagen, [email protected], 2015-03-09
Glossary II/IIIBeam Production Chains & Patterns
● Beam-Production-Chain:– organisational structure to manage
parallel operation and beam transfer through FAIR accelerator facility
– defines sequence and parameters of beam line from the ion-source up to an experimental cave (e.g. APPA, CBM, SuperFRS, ...)
– definition of target beam parameters (set values): isotope, energy, charge, peak intensity, slow/fast extraction, …
● Beam Pattern:– grouping of beam production-chains
that are executed periodically
– can be changed of pattern within few minutes (target, requires automation for beam-based retuning)
ChainsSIS18
SIS100
HESR
SIS18
SIS100
HESR
Pattern
SIS18
SIS100
HESR
courtesy D. Ondreka
GSI Helmholtzzentrum für Schwerionenforschung GmbH 47Ralph J. Steinhagen, [email protected], 2015-03-09
SIS18SIS100
Unilac
CBM + RIB ext. target (U73+) + AP (LE)
SIS18SIS100
Unilac
ESR
RIB ext. target (U28+) + ESR
AP + RIB ext. target (U28+) + Biomat
SIS18SIS100HESR
Unilac
SIS18SIS100
Unilac
ESR
RIB ext. target (U28+) + ESR
SIS18SIS100
Unilac
CBM + RIB ext. target (U73+) + AP (LE)
AP + RIB ext. target (U28+) + Biomat
SIS18SIS100HESR
Unilac
Periodic beam patterns, dominated by one main experiment:
Glossary III/IIIExample: Chain & Pattern
courtesy D. Ondreka
GSI Helmholtzzentrum für Schwerionenforschung GmbH 48Ralph J. Steinhagen, [email protected], 2015-02-24
Terminology: Accuracy & Precision (also ISO 5725)
Good summary: http://en.wikipedia.org/wiki/Accuracy_and_precision
● Accuracy: “[..] closeness of measurements [..] to its actual (true) value”
● Precision (also: reproducibility or repeatability): “[..] degree to which repeated measurements
under unchanged conditions show the same results.”
● Example: “Target analogy” and the two extreme cases
● Resolution: smallest change that produces a response in the measurement
High accuracy, but low precision High precision, but low accuracy
we need this from the BPMsobtained through beam-based alignment
GSI Helmholtzzentrum für Schwerionenforschung GmbH 49Ralph J. Steinhagen, [email protected], 2015-03-09
FAIR Challenges & Constraints… SIS18 Operation Experience & Efficiency I/II
● possibly pessimistic/simplistic1,2 estimate, control room experience:
– presently: '~ 1 shift UNILAC setup + 1 shift SIS18+TL setup' ↔ 1-2 weeks of experiments
● 2014: Beam-on-Target (BoT) figure of merit (FoM) of ~70%
– sufficient for present mode of operation (~20% HW failures, ~13% setup)
– however: high losses/activation & FoM does not scale for FAIR
1possibly strong assumption that new machines can be operated with the same routine, ease and efficiency as the present GSI infrastructure, …2complex beam chains (e.g. HESR) with long beam setup times are typically run longer/more static than shorter (SIS18 experiments)
GSI Helmholtzzentrum für Schwerionenforschung GmbH 50Ralph J. Steinhagen, [email protected], 2015-03-09
FAIR Challenges & Constraints… SIS18 Operation Experience & Efficiency II/II
“… room to improve!”
εFAIR := ∏i
nmachines
εi = εESR/SIS18⋅εSIS100⋅εSuperFRS⋅εCR⋅εHESR⋅...
● FAIR-BoT (efficiency εFAIR
):
– further convoluted with HW failures, availability of infrastructure
GSI Helmholtzzentrum für Schwerionenforschung GmbH 51Ralph J. Steinhagen, [email protected], 2015-XX-YY
SIS100 → p-Target Steering Analysisusing all BI Instruments and Quadrupoles
J.Pietraszko, HIC4FAIR Workshop, Darmstadt, February 26 ,2016
Beam position instability – consequences for pion beam
- maximal SIS intensity (N beam), - Be target diameter 4 mm loses if the beam is not stableplan to install Be target of 2.3 mm diameter !